x : \ (t ,4 4'"; \g c. f;- .. h J.» o . u .a u . . u 4 C . .n .s O L ”1” O . ‘ This is to certify that the thesis entitled "A Consideration of Weed Control Through Physical Properties of Seeds" presented by William E. Splinter has been accepted towards fulfillment of the requirements for ____M- S 0 degree inéfii‘WEI.ral Engineering UWMGALR Major prbfessor Date December 4, 1951 0-169 A CONSIDERATION OF WEED CONTROL THROUGH PHYSICAL PROPERTIES OF SEEDS By William Eldon Splinter WM“- AN ABSTRACT Submitted to the School of Graduate Studies or Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Engineering 1951 268357 A CONSIDERATION OF WEED CONTROL THROUGH PHYSICAL PROPERTIES OF SEEDS The problem of weeds in sugar beet fields is one of the major problems of the industry today. Even the best of present day methods of weed control requires at least one hoeing of the field by hand. A possible solution of the problem would be killing the weeds in a narrow strip of soil by heating them with.an alternating electric field. If there is enough difference in dielectric preperties of soil and seeds, and between seed species themselves, selective heating may be possible. To determine the feasibility of selective dielectric heating an investigation of the variation of dielectric con- stant and power factor of seed components with frequency and moisture content was initiated. For this preliminary work the components of wheat were used because they are available at a high degree of purity. A significant variation of power absorbing ability between the components of wheat and variation of power absorbing ability with frequency is evidenced by the results of this work. This power absorbing ability or heating rate increases with frequency and moisture content for all of the components, but at different rates. A CONSIDERATION OF WEED CONTROL THROUGH PHYSICAL PROPERTIES OF SEEDS By William.Eldon Splinter A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Engineering 1951 ACKNOWLEDGMENTS The author is grateful to have this opportunity to ex- press his thanks to those who have made this work possible through their advice and contributions. First of all, thanks must be given to Professor A. W. Farrell, for it was through his efforts that the Opportunity to take advanced work at Michigan State College was made possible. The author is deeply indebted to Dr. W. M. Carleton for his continual guidance and many helpful suggestions. Thanks must be given to Professor B. R. Churchill of the Farm Creps Department for his extensive and very valuable assistance on the problem. The cooperation and advice of Professor I. O. Ebert of the Electrical Engineering Department is deeply appreciated, and the advice given by Dr. R. D. Spence of the Physics Department on measurement theory is gratefully acknowledged. The Farmers and Manufacturers Beet Sugar Association, Mr. A. A. Sohupp and Mr. Perc Reeves in particu- lar, are to be thanked for the fellowship under which this work was done. Finally, the author would like to express his thanks to the Agricultural Engineering, Farm Crops, and Electrical Engineering Departments as a whole for the use of their equipment and laboratories for the experiments. 11 TABLE OF CONTENTS Page INTRODUCTION ------------------------------------ ----- 1 REVIEW OF LITERATURE --------------------------------- Weed Control Methods ............................ 3 5 Dielectric Heating .............................. 5 THEORETICAL INVESTIGATION OF DIELECTRIC HEATING ------ 6 Effect of Heat on Life -- ........................ 5 Heating Effect of Radio Frequency Field -------- - 8 Variation of Dielectric Constant and Power Factor with Frequency ------------------ - .......... 9 Variation of Dielectric Constant with Temperature 10 Variation of Dielectric Constant with Moisture -- lO Relation between Dielectric Constant and Dipole Moment ------------------- ----------;------- 11 Dielectric Constant of Soil --------------------- 12 THEORY OF MEASUREMENT -- .............................. 13 EXPERIMENTAL INVESTIGATION ------ .................. --- 15 Materials Used -- -------------------------- - ..... 15 Apparatus and Methods ------------------ - ..... --- 16 Experimental Results - ........................... 25 Interpretation of Results ---- ------- - ........... 56 DISCUSSION ---------- - ---------------------- ---------- 38 BIBLIOGRAPHY ------------------------- - ............ --- 41 iii LIST OF FIGURES Page Figure 1. Variation of dielectric constant (e) and power factor (tan 9) with frequency ------ 10 Figure 2. Schematic diagram of equivalent test circuit ------------------------------- --- 16a Figure 5. Schematic diagram of test circuit at high frequencies --- -------------------- -- 16a Figure 4. The Boonton 160A "Q" Meter with empty test condenser and inductance connected -- 17 Figure 5. The 45 upF Condenser -- .............. ----- 21 Figure 6. The 58 ppF Condenser --- ------------------ 21 Figure 7. Disassembled 58 ppF Condenser ------ ------ 22 Figure 8. Variable inductance constructed from an antenna tuning coil ----- ----- - -------- 25 Figure 9. Inductance coils which were connected in series with the condenser for resonance at various frequencies -- ------- 25 Figure 10. Ennidity Control Chamber -------- ........ - 24 Figure 11. The variation of dielectric constant of wheat endosperm with frequency and moisture content --------- ................ 25 Figure Figure Figure Figure Figure Figure 12. 13. 14. 15. 16. 17. iv Page The variation of power factor of wheat endosperm with frequency and moisture content ------e --------------- 28 The variation of dielectric constant of wheat bran with frequency and moisture content -- -------- --- --------- 29 The variation of the power factor of wheat bran with frequency and moisture content - --------------------- 31 The variation of dielectric constant and power factor of wheat germ with frequency --- ------------- - ------ - ----- 32 The variation of relative heating rate of grain components with frequency for any given voltage -— --------------- ---- 54 The relative power absorption or heating rate of wheat components as percent of total power input ----. ............. --- 35 INTRODUCTION In the past, the raising of sugar beets has always been closely associated with hand labor. All of the field opera- tions, blocking, thinning, weeding, tepping and loading, have been tedious manual operations. Within recent years, however, many advances have been made in the mechanization of the raising of sugar beets. Improved tillage methods, planting with segmented and decorticated seed, mechanical stand reduction and mechanical harvesting have all contri- buted to the reduction of hand labor. Although many of these developments are still in the experimental stage, it seems probable that most of the hand labor can soon be replaced by machine methods. There is, as yet, one major stumbling block before the complete elimination of hand labor from beet fields-~the problem.of weeds. The complicated nature of dormancy, along with the very long life of weed seeds adds to the difficulty of weed control. Present day methods of hand blocking and thinning re- quire many hours of tedious work with hoes. Even with the newer methods of mechanical thinning and stand reduction, the Operation must be followed by at least one hoeing. Results from the field indicated that if weeds could be controlled in a narrow strip on either side of the row, the major part of the problem would be solved. Weeds could be easily handled between the rows with present day cultivation methods. One possible method of eliminating weeds from this narrow strip would be through heating. However, to heat the soil and accompanying seeds with flame or an electric current would require a great deal of energy and probably result in excess- ive drying of the soil. A possible solution might be selec- tive instantaneous heating by an alternating electric field. If the physical prOperties of the soil and seed constituents are such that the power absorption of one material is higher than another at a given frequency, then selective heating may be possible. For a fundamental approach to the feasibility of weed control through dielectric heating, the dielectric constant and power factor of seeds and seed components (germ, endo- sperm and hull) must first be evaluated, therefore, an inves- tigation of the variation of the dielectric constant and power factor with frequency and moisture content of seed components has been attempted. For this initial investiga- tion the germ, bran and endosperm of wheat have been used. REVIEW OF LITERATURE Weed Control Methods Chemicals have assumed an important role in the control of weeds within recent years. The non-selective types (chlorates, arsenates, borates, etc.) have been used for several years in the control of heavy infestations of weeds in small areas (4, 41, 42, 43, 46). The recent deve10pment of selective herbicides, however, has put a new emphasis on chemical farming. Probably the most widely known of these selective herbicides is 2, 4eD. A large amount of research work has been carried on with this chemical (4, 22, 24, 26, 27, 46), and its use is now quite generally accepted in the control of broad leaved plants. Another group of chemicals, the trichloroacetates, has been found to control grasses, leaving broad leaved plants relatively untouched (50). Recent work by Anderson and Lyerly (l) in Texas and by Talley (46) in Mississippi has indicated that control of many weeds and grasses by herbicidal oils is practical. Certain oils have been used as a selective spray in the cotton fields with good results. Pigweeds under three inches tall have been killed with one dosage. Flame cultivation has also been used extensively in the cotton fields. Fairly good weed control was obtained, although Danielson and Grows (15) stated that pigweeds and a few other weeds and grasses survived the flaming. An investigation of the control of weeds by mechanical means by McBirney (29) showed that from 22.2 percent to 34.9 percent of the weeds could be removed by various implements. The miked tooth harrow had the highest reduction in weeds. Soil pasteurization has been practiced for a number of years in greenhouses. This sterilization of the soil not only kills weed seeds, but also parasites, nematodes, and insects. Newhall (32) found that the electrical energy required to raise the temperature of one cubic foot of soils from.15°C (59°F) to 65°C (149°F) for pasteurization was: TABLE 1 ELECTRICAL ENERGY FOR HEATING SOIL Soil KWH Mean Type ‘ Per Ft3 Watts oC Sand 1.15-1.68 27 Loam. 1.16-1.62 28 Muck 1.54-2.05 55 Lachman (24) found that Stoddard Solvent could be used successfully as a pre-emergence spray for garden beets. The use of ordinary sodium chloride sprays as a selective herbicide on sugar beets has been found to have some merit. Nichol (55) found that ragweed, wild mustard, pigweed, smart- weed and small annual grasses could be controlled but no con- trol was obtained over lambsquarter, sow thistle and quack- grass. Other workers (51) found that aromatic oil-pentachlo- rcphenol sprays controlled the weeds with.no damage to the sugar beets in an experimental plot. Dielectric Heating Bitter (8) placed some barley seeds on the lower conden- ser plate of a short wave oscillator circuit. He tried various frequencies and exposure times but obtained no effect on the germination of the seed. Bellinzaghi (6), however, found that inhibition of .212$§.££EE (English dwarf beans) results after a fifteen to twenty second exposure at a frequency of one hundred and seven megacycles. Using a frequency of fifteen megacycles and exposures of two, three, four, and five minutes, Lambert, et a1 (25)" treated wheat, oats, barley, perennial peppergrass, Canada thistle, field bindweed, wild mustard, wild oats, quackgrass and leafy spurge. ' With a two minute exposure they reduced the germination of wild mustard from 22 percent to 5 percent and the germina- tion of quackgrass from.88 percent to 9 percent. THEORETICAL INVESTIGATION OF DIELECTRIC HEATING Effect of Beat on Life The exposure of seeds to a radio frequency field will, to some extent, alter the germination of the seeds. One of the possible effects of this alternating electric field is that of dielectric heating, with a corresponding increase in temperature. ’ This increase in temperature has a very definite effect on the life process of the seed. A Chemical reactions which have a measurable rate of re- action show an increase in rate of reaction with an increase in temperature. This increase in rate of reaction can be expressed theoretically by the van t'Hoff-Arrhenius law of physical chemdstry. h7fié..A(E-I ' ‘K'E or, _ 53-. = eAE‘Tfi-I') K. where, K2 is the equilibrium constant of the reaction at a temperature T2, K1 is the equilibrium constant of the reaction at a temperature T1, and, A is the temperature characteristic. This ratio —%fi is often expressed as a temperature coefficient ”Q". For many life functions the value of Q for a temperature difference of 10°C ranges between two and three. This means that, with an increase of 10°C in temperature, the rate of reaction increases two or three fold. According to Rahn (58) the value of Q for’many proteins ranges between ten and one hundred for a 10° temperature rise. The heat coagulation of the protein Hemoglobin which takes place in twenty-six seconds at 80°, takes six minutes at 70°, ninety minutes at 60°, 20.6 hours at 50° and twelve days at 40°. If one of the vital proteins essential for the life process of the cell coagulates completely, the cell is dead. Also, most of the cell catalysts are similar to, or are linked to, native proteins. It has been found that the thermal death- point for a life process is fairly definite, being only a few degrees above the maximal temperature. The maximal tempera- ture is that temperature at which the process proceeds at its maximum rate. This indicates that it may be possible to kill the seed germ by an almost instantaneous application of heat if the preper temperature is attained. Since the nature of the pro- teins in the various seeds differs, and since the various pro- teins do not have the same coagulation temperature, it may be possible to obtain selective killing of different species by the proper temperature control. Heating Effect of a Radio Frequency Field A dielectric, placed in an electrostatic field, is sub- Jected to two forces: (1) the distortion of the molecules of the dielectric in which the orbits of the negatively charged electrons are displaced in the direction of the positive charge of the field and the positively charged nucleus is displaced, to a very small extent, in the direction of the negative charge of the field; and (2) the rotation of the molecules due to their dipole moment. If the electrostatic field is alternated very rapidly, the internal resistance to this distortion causes a heating of the dielectric. According to Terman (47), this heating effect may be ex- pressed by: H (0.55 x10'12) (3% (E2) to) Lari: watts/cc. d where, f is the frequency in cycles per second, E is the difference in potential between the condenser plates, 1 p.f. is the power factor of the material, e is the dielectric constant, and d is the distance between the condenser plates in centimeters. It can be seen from this equation that for a given volt- age gradient and plate distance, the heating effect of a mate- rial in an electric field varies directly with frequency, dielectric constant and power factor. Variation of Dielectric Constant and Power Factor with Frequency When a dielectric material is subjected to an alternat- ing electric field the polarization of the dielectric will depend on the frequency and will be out of phase with the field. At low frequencies the ions and dipoles will contri- bute to their fullest extent to the dielectric constant with- out an appreciable time lag. At higher frequencies the motion of these ions and dipoles can not keep up with the change in field direction and will therefore contribute less to the dielectric constant. Therefore, the dielectric constant will decrease with increase in frequency. The ions and dipoles get more and more out of phase as the frequency increases. This phase difference (0) is the loss angle and tan 9 is the power factor. This phase lag gives rise to heat dissipation which is therefore preportional to the frequency and tan 0. Where the frequency corresponds to the relaxation time of the dipole, tan 9 becomes a maximum. The relationship between the dielectric constant and the power factor can be seen in Figure 1. -10- Variation of Dielectric Constant with Temperature Morgan (51) has found that, in some dielectrics, there is evidence of a cooperative effect such that when one mole- cule starts to rotate it makes it easier for its neighbors to do so, and the effect spreads. These materials exhibit a sharp transition zone in which, with.but a few degrees change in temperature, the dielectric constant increases three to eight times its original value. Variation of Dielectric Constant with.floisture Berliner and Ruter (7) obtained values of the dielectric constants of wheat and rye at various moisture contents. The values ranged from.thirty at 10 percent to forty-three at 16 percent moisture and sixty at 19 percent moisture. These values indicate that water,dielectric constant 81, has a very large -11- influence on values of the dielectric constant of grain. The dielectric constant of starch and gluten (principal components of the endosperm) ranges between three and four. Relation Between Dielectric Constant and Dipole Moment From.the Clausius-Mosotti equation of physical chemistry, .(-1) (M P T‘ET‘TH a. where, P'is the molar polarization of the substance, e'is the dielectric constant, M is the molecular weight, and d is the density, one can obtain, e= ale—'3“— _c1...p which indicates, that, with an increase in molar polarization there is considerable increase in dielectric constant. Hewever, P = PD-Fu where Pb is the induced or distortion polarization and Ba is the permanent polarization. According to Debye, 47”,“: .. Alec = ’3 ‘ 47 a"! 9‘ 9KT where N is Avogadro's constant (6.02 x 1023), a is the polarizability of the dielectric, ,u is the dipole moment, and K is the Boltzmann constant (1.58 x 10‘15), -12- It can be seen that, for unsymmetrical molecules, the dielectric constant increases with the square of the dipole moment. Dielectric Constant of Soil Ratcliffe and White (59) found a decrease in dielectric constant with an increase in frequency for soil. The values decreased from.about forty at .25 megacycles to eleven or twelve at frequencies above three megacycles. Banerjee and Joshi (5) found the dielectric constant of soil to be approximately three at six percent moisture at a frequency of seventy megacycles, increasing to 12.5 at four- teen percent moisture. They also found a decrease in dielectric constant with an increase in frequency. THEORY OF MEASUREMENT Measurement of dielectric constant and power factor at frequencies ranging from 50 KC to 40 MC were accomplished with the aid of a Boonton 160A "Q" Meter. I By definition, Q is the ratio of the reactance to the resistance of the reactive element in question. This relation- ship may be expressed by: XL , 27/1 Q-7r - ’I for an inductance, and 3 X - 1 Q 7'?“ “27“,? for a capacitance where XL is the inductive reactance in ohms, Xe is the capacitive reactance in ohms, R is the resistance in ohms, f is the frequency in cycles per second, L is the inductance in henries, and, C is the capacitance in farads. For a simple series circuit at resonance, X ='xc and the L current will be the same through all components. Therefore -14- where E is the total voltage across the circuit, E; is the voltage drop across the condenser, Z0 is the impedence of the condenser, and, Re is the effective series resistance of the circuit. Since the reactance is of considerably greater magnitude than the resistance of the condenser, E,=X:Q:X h c w ere Q is the effective Q of the E R: e (Hg, circuits In the "Q" meter Ec is read by means of a voltmeter having negligible power consumption and E is held constant, therefore Q can be read directly from the voltmeter. For inductances the true Q differs from Q. This differ- ence is due to the distributed capacitance Cd of the coil and can be expressed very closely by Q: (1+Cd). Q9 '6' Since the power factor is, by definition, the ratio of the resistance to impedance, therefore, Pfi‘W‘ ‘2: For impedences having Q greater than ten this is correct within one percent. . v. I w . . . ,. 1 U \ ‘ a . I I ' I ', a , , U n -15- EXPERIMENTAL INVESTIGATION Materials Used An investigation was started to determine the effects of dielectric heating on seeds. Since the protein content of the seed germ is considerably higher than that of the endo- sperm or of the hull, it was felt that there might be enough difference in dielectric constant to allow for differential heating at radio frequencies. Preliminary experiments were started to determine the difference in dielectric constant of the components of seeds. For this work wheat was chosen because the components of the seed are available in quantities and at a high degree of purity. Unbleached flour was obtained from the King Milling Company of Lowell, Michigan. Finely ground wheat bran was obtained from the National Biscuit Company of Cleveland, Ohio, and the wheat germ meal of over 90 percent purity was obtained from the Pillsbury Mills, Incorporated of Minneapolis, Min- nesota. A The average percentage of protein and density of these ' components are: -15- TABLE 2 PERCENTAGE OF PROTEIN AND DENSITY OF MATERIALS Protein Content Percentage Densityg Wheat flour 8.55 0.58 g/cc. Wheat bran 15.51 0.59 g/cc. Wheat germ 55.46 . 0.62 g/cc. Determination of the composition of the materials used was made by the Agricultural Chemistry Department. The density was determined in a packed condition as nearly similar to that in the condenser as was possible. Apparatus and Methods The method used in the determination of the dielectric constant and power factor was the insertion of the test condenser in parallel with the calibrated condenser of the "QP meter and resonating this capacitance with the various inductances. . The schematic diagram of this circuit is shown in Figure 2. -163- L R %———I I I ____ .__.L.._. C ' c, ( —o————l Figure 2. Schematic diagram of test circuit at low frequencies. L R 2 5 ‘9—--—-' c»: C>5L ( x _ -..L._.. N __T__ M c ,1 c, ’\ Rs _ 9—-——-J 5 4 Figure 5. Schematic diagram of equivalent test circuit at high frequencies. Meaning of Symbols L is the external inductance used to resonate the circuit. R is the series resistance of the circuit. M is the voltmeter from which Q is read. C is the calibrated variable capacitance. Cxis the external test condenser. ins the lead inductance of the test condenser, and R,is the series resistance of the test condenser. -17- Readings of frequency, Q1 and C1 were taken. The test condenser was removed, the circuit re-resonated at the same frequency, and readings Q2 and C2 taken. Iith.these readings the dielectric constant and power factor can be calculated as follows: Since C - C a C = the capacitance of the test 2 1 c condenser, then a 3 4.45.C£'d where d is the distance between the condenser plates in inches, and, A is the active area of the condenser plates in square inches. Ce Q! 03 C3 (Q3 ‘Jo‘ = -—--—" A a .. 9 ' ’1’“ Q c.(o.-o.) ’ t "M" ’f 6.0.0. Figure 4. The Boonton 160A "Q" Meter with empty test condenser and inductance connected. At high frequencies there was an apparent increase in the capacitance of the condenser due to the inductance of the condenser leads and plates. The value of this lead inductance was determined by clamping the condenser plates t0gether, there- by shorting them out, and testing the condenser as an induct- ance. 0n the theory that there may have been some capacitance in the system due to oxide coating or poor contact of the plates, the distributed capacitance was determined but was found to be only one percent of the total capacitance of the circuit, and therefore neglibible. The inductance of the condenser leads was found to be 0.26 f¢H, which was insignificant at low frequencies but very impor- tant at high frequencies where the inductance of the resonating coil may be as low as 0.08/(H. The resulting circuit at high frequencieswas no longer a simple series circuit, but changed to a combination series- parallel circuit as shown in Figure 5. The Q of this circuit can not be used to calculate the Q of the condenser and 02 - C1 does not give the actual capacitance of the test condenser. Calculations of capacitance and power factor were made as follows: In the case of the series circuit, the total impedance of the circuit is equal to the sum of the impedance of the inductance and the impedance of the capacitance, or 2,: Z, + Z. = V R“2 + (Xc - sz. At resonance the inductive -19- reactance equals the capacitive reactance (Xc : XL') therefore Z, is simply the series d.c. resistance of the circuit. Hewever, referring to Figure 5 for the combination series-parallel circuit, 215' 214 Z 2 Z + t Z ’2 2" #234 Since 212 is primarily inductive, the distributed capaci- tance and resistance of the coil can be neglected and 212 can be represented by XL. Neglecting inductive reactance and resistance of the variable capacitance, Z can be represented 25 by X0. The total impedance then becomes, Xe 2.74 Xg {—234 For the system.to resonate, some equivalent capacitive z.=x,_+ series reactance must equal the inductive reactance of XL. The value of this reactance was determined by resonating the circuit at the same frequency used for'testing, but without the test condenser connected. The reading of the calibrated condenser then gave the value of the capacitance for computing the equivalent capacitive reactance Xét Another method of calculating this reactance would be the determination of XL itself but it was felt the capacitive reading would be less subject to errors brought in by the flux linkage and distri- buted capacitance of an external coil. The capacitive reactance of the test condenser was determined by solving the following equation for Xéz = x, 2,, _ x, /R,‘ +(x,”-.\1’)‘ - *— xcl "‘ +2.. X. + 70:: + cxgzxxr where X4 is the equivalent capacitive reactance, X0 is the reactance of the variable condenser, Kg' is the capacitive reactance of the test condenser, Xi is the inductive reactance of the test condenser, and R3 is the effective series resistance of the test condenser. ' ’ 1 Therefore, Xc" = X[ :1:- 7M) __ R} X;*XJ from which the positive root was used. The capacitance of the test condenser can then be deter- mined from.C : 332527 The power factor was determined from the ratio, p.f. 3 go where Z.‘W . The capacitance of the leads of the condenser was determined by testing the capaci- tance of similar leads without plates attached. This value was added to the value of 02. The increase in the distance between the plates of the test condenser due to the material being tested was determined with a micrometer and the value added to d in the determination of the dielectric constant. Three condensers were constructed and used. The 45 ppf condenser (Figure 5) was constructed of wood with four inch square sheet steel plates. The ends of the condenser were removable for filling and cleaning. This condenser had too much inherent inductance and was discarded. Figure 5. The 45‘upF Condenser. This condenser is constructed of wood with four inch square sheet iron condenser plates. The ends of the condenser are removable for ease in filling and emptying. The 58 uuF condenser (Figures 6 and 7) was constructed of Plexiglass with three inch square aluminum plates. Connect- ing leads were of stiff silver wire and all bolts were of brass. This condenser was designed so that one side of the condenser was removable, making filling and cleaning easier. The 54 upF condenser was similar to the 58 ppF condenser but the distance between condenser plates was reduced from 0.056 to 0.040 inches. Figaro 6. The 58 ppF Conaerxar. This condenser is constructed of Plexiglass with three inch square aluminum condenser plates and silver wire leads. é. -22- Figure 7. Disassembled 58 uuF Condenser. This improved design with one side removable allowed for easier filling and emptying. Corrections for fringing were made using the formula: _ 75rd 2-5 {-4’ 7574-1) C‘- — 0.07/6 L,[ d f /n(——————d + /n d ]; where C518 the capacitance of the condenser with fringing, L is the length in inches, b is the width in inches, and d is the distance between plates in inches. A variable inductance (Figure 8) of from zero to sixty turns was constructed from an antenna tuning coil but it was found to be unsatisfactory at high frequencies. Separate inductances (Figure 8) were connected in series with the condensers for resonance at the various frequencies. This variable inductance was constructed from an antenna tuning coil. It proved unsatisfactory at high frequencies because of harmonics. ' Figure 8. Figure 9. Inductance coils which.were connected in series with the condenser for resonance at various frequencies. -24- The water content of the material was regulated by ' constant humidity jars (Figure 10) containing solutions of sulfuric acid and distilled water. The material was suspen- ded in containers above the solution and the partial pressure of the water vapor, as regulated by the strength of the acid solution, controlled the moisture content. Figure 10. Humidity Control Chamber. The moisture content of the material is regulated by suspend- ing the material above a given solution of sulfuric acid and water. The percentage of moisture was determined by heating a weighed sample of the material in a constant temperature air oven at 150°C for one hour and reweighing after cooling in a dessicator. The percentage of moisture was determined on a wet basis. This method, though not the most exact, was found by Burton (10) to be fairly consistent with results obtained by the standard method of slow heating in a vacuum.oven at 98-1000 0 for forty-eight hours, and with.the Brown-Duval method. For the purpose of these preliminary tests this approximate method was considered sufficient. Experimental Results The variation of the dielectric constant of wheat endo- sperm with frequency and moisture content can be seen in Figure 11. There is evidence of the Debye resonance at frequencies above nine or ten megacycles. The upper limit of this resonance, where the molecules are rotating in the same period as the frequency of the field, was not determined. The relatively slow decrease in dielectric constant indicates that there is a range in the size of molecules and, as the frequency increases, less and less of the molecules are able to rotate with the changing field, thereby adding less to the dielectric constant. There is some indication that there may be another Debye resonance at frequencies below fifty kilocycles, which.was the lower limit of the Q,meter. The value of the dielectric constant decreased approximately thirty percent in the range of frequencies from fifty kilocycles to nineteen megacycles. An increase in moisture content adds considerably to the dielectric constant but without an appreciable alternation of the basic frequqncy curve. .paouooo oedemaos one hesosooam and: eacamooco uses: 90 pompmcoo cappeoaoac no codename» age .HH oaswum A mmfl 0.x. SEES W ozmbdmmm ma 0H ma NH OH 0 w m N o R a u F O o.n m.n . e.» ZKEAmOCZm B qqusuoo oyaqoetetq The variation of the power factor of wheat endosperm.with frequency and moisture content is shown in Figure 12. There appears to be a levelling off of the power factor at fre- quencies above twelve megacycles, but a further increase is noted at around nineteen megacycles for the endosperm at 9.1 percent moisture. This increase indicates that there might be a levelling off of the dielectric constant, above nineteen megacycles, possibly in the approach to another Debye resonance at some higher frequency. In the frequency range used (0.05 to twenty megacycles) therewas an approximate 100 percent increase in power factor for the endosperm.at 7.5 percent moisture and 140 percent increase at 9.1 percent moisture. There is an apparent increase in power factor with an increase in moisture content although the power factor for 9.1 percent moisture and for 12.8 percent moisture were practically the same. The variation of dielectric constant of bran with frequency and moisture content is shown in Figure 15. A Debye resonance is found, beginning at frequencies above twelve megacycles. The decrease in dielectric constant is gradual, indicating a considerable variation in molecule size. The dielectric constant of the bran at 6.6 percent moisture approaches a value of one, which is the value of the dielectric constant of air. An increase in moisture content served only to increase the dielectric constant without significantly altering the shape of the curve. -28- .ufiopcoo manumaofi can honozooam flows anomoeno each: «0 gonowu aceoa mo coapeaamp one Ammqofio¢umzv Mozmbdmmk .ma enemas 02 NH 0H «H NH OH m _ a u T _ d Emma moazm 9553 to moeo 0mm mm.e. .4 3¥—-c> OH HH NH queqsuoo oxaqoeterq -29- .uoopooe enoumfioa can honoaooah nude scan been; mo pseumooo oaauooaoap mo Sofiumaas> ssh .nH oazmfia AmquMo L L 0mm meme . . 0mm mm.0H. O.H N.H ¢.H 0.H m.H O.N N.N ¢.N queqsuoo oyaqoetetq -30- The power factor of the bran (Figure 14) levelled off at about ten megacycles and gradually declined with an increase in frequency. The values for the dielectric constant and power factor of bran are considerably less than those for the endosperm. The variation of dielectric constant and power factor of wheat germ are shown in Figure 15. There is apparently a Debye resonance between ten and eighteen megacycles, evidenced by a decrease in dielectric constant and maximum.power factor. The power factor and dielectric constant vary abruptly indicating a fairly uniform molecule size at this point of resonance. There is another drop in dielectric constant and power factor at frequencies between twenty-four and twenty-eight megacycles. The exact nature of this drop was not determined but the abruptness of the dr0p would indicate another resonance point of molecules fairly uniform in size. A comparison of the three components shows that the beginning of at least one Debye resonance occurs for all of the materials at a frequency of about ten megacycles. For the endosperm.the decrease in dielectric constant continued up to twenty megacycles but the power factor curve indicates that there may be a levelling out of dielectric constant and another Debye resonance at some higher frequency. This would be similar in behavior to the germ. There is also an indication that another resonance may be found for frequencies below fifty -31- .pCcpsoo caspmaos one hocosvcpm sud: caps ummfii MO houomh LQBOQ 03» HO Gofipwfihmb QSE .VH ohfiwfih Ammquodomzv Mozmbdmmm ow mm on mm om ma 0H 0 - - I - - - q a L ////If . 0mm wo.o Z¢mm E omm “n.0- l L 4 qusqsuog otaqoetexq .bocoSUopa no“: snow uses: ho honoah 90309 was peduncOo oakuooflofic no ceauwfina> 0:9 .mH ohsmak Ammuofio pcmumcoo ofiauoeaowa pouomm pekom . m.w o.¢ m.v Joqoeg Jemog mm.m om.m mb.m 00.0 queqsuog oyaqoeterq -53.. kilocycles. The bran exhibited a gradual decline in dielectric constant and power factor with no abrupt changes in value. The germ evidenced the most pronounced Debye effect of the three, two very definite points of resonance being noted. The relative rate of heating of an alternating electric field for the three components is shown in Figure 16. This value was computed from the frequency, dielectric constant and power factor, and serves as a basis of comparison of the relative heating rates that could be expected using a given voltage and distance between condenser plates. The power absorption of the endosperm and bran varies uniformly with the frequency but two pronounced plateaus are noticed for the germ. I The heating rate of the endosperm.and germ increases very rapidly with an increase in frequency. With any given voltage, doubling the frequency very nearly doubles the heating rate or power absorption of the material. At the two plateaus in the curve for the germ, however, an increase in frequency does not increase the power absorption. The heating rate of bran varies slowly with frequency and reaches a maximum at around thirty-five megacycles. The relative power absorption of the three components, expressed as percentage of the total power absorption is shown in Figure 17. Within the frequency range of 0.05 to twenty megacycles the power absorption of the bran remains almost constant between eight and ten percent. The power absorption -34- .owmpHo> no>am has now hocosaohh and: mucoCOQEoo namaw no cash mcfipwon o>aumdoh no neausaaa> .oH madman A waowudomzv HozmSGmE 0v mm on mm ON ma Ofl n O Nz< ho OZHBH5 zm>Hu GOR mezmzomzoo Bdmmg own mmé Shonmoccm eqeg Buyqsea catastea .usmcfi aokom kuOp ho poooaom mm mucczoasoo pecan go open wcapmon ao soapQAOmnm peace o>HpaHoa one .bH madman AmMAoNOH9